Adulteration is frequently encountered in the food industry and can be identified using currently available techniques. Infrared and Raman spectroscopic procedures are the most attractive techniques regarding fats and oils. The objective of this study was to determine the adulteration of the fat source (margarine or butter) in bakery products using Raman and near-infrared (NIR) spectroscopies. Margarine and butter samples were purchased at local markets in Turkey and examined using Raman and NIR devices. A mixture (50 % margarine : 50 % butter) of fat samples was examined as well. The NIR and Raman spectral output data of all the fat samples were processed using principal component analysis (PCA). Good classification was obtained for margarine, butter and the 1:1 adulterated mixture. The chosen bakery product (cake) was produced using the same fat samples according to the method of the American Association of Cereal Chemists. Then, the fat fraction was extracted from the cakes with n-hexane. Extracted fat samples from the cakes were examined as before. PCA was applied to Raman and NIR spectral data to achieve the separation of fat sources in the cakes. PCA was also validated in each of the two stages. Significant decomposition was observed in the Raman study in contrast to the NIR study. A chemometric comparison was also applied to processed (baked) fat samples in cakes and purchased samples by PCA to assess the effects of heat treatment on sample spectra. Raman spectroscopy with multivariate analyses such as PCA can be used to detect the adulteration of the fat source in bakery products in a faster and more suitable way than the other methods.